1153

Inflamm Bowel Dis • Volume 27, Number 7, July 2021

Received for publications September 10, 2020; Editorial Decision November 10,

2020.

From the *Department of Biomedicine, Unity of Pharmacology and

Therapeutics, Faculty of Medicine of the University of Porto, Porto, Portugal;

†Department of Internal Medicine, Tâmega e Sousa Hospital Center, Padre

Américo Hospital, Penafiel, Portugal; ^‡Department of Gastroenterology, Vila Nova

de Gaia/Espinho Hospital Center, Vila Nova de Gaia, Portugal; ^§Unit of Clinical

Pharmacology, São João Hospital Center, Porto, Portugal

Author Contribution: FJM, PPL, and MME contributed to the study concept and

design, acquisition of data, and analysis and interpretation of data. FM contributed to the

study concept and design, acquisition of data, analysis and interpretation of data, study

supervision, and critical revision of the manuscript for important intellectual content. All

authors read and approved the final version of the manuscript, including the author list.

Supported by: This work was supported by Grupo de Estudo da Doença

Inflamatória Intestinal (GEDII). FJM is supported by funding from Fondazione

Cariplo via the “Recruiting and Training Physicians-Scientists to Empower

Translational Research: A  Multilevel Transdisciplinary Approach Focused on

Methodology, Ethics and Integrity in Biomedical Research” project (project grant:

02.00280, clinician scientist project).

Conficts of Interest: FM served as speaker and received honoraria from Merck

Sharp & Dohme, Abbvie, Vifor, Falk, Laboratórios Vitoria, Ferring, Hospira and

Biogen. All other authors have nothing to declare.

Address correspondence to: Fernando Magro, MD, PhD, Department of

Biomedicine, Unity of Pharmacology and Therapeutics, Faculty of Medicine of

the University of Porto, Rua Plácido da Costa, 4200-450 Porto, Portugal, E-mail:

fm@med.up.pt.

© The Author(s) 2020. Published by Oxford University Press on behalf of

Crohn’s & Colitis Foundation. All rights reserved. For permissions, please e-mail:

journals.permissions@oup.com

Basic science Review aRticle

The Role of Dipeptidyl Peptidase 4 as a Therapeutic Target

and Serum Biomarker in Inflammatory Bowel Disease:

A Systematic Review

Francisco Jorge Melo, MD,* Pedro Pinto-Lopes, MD,*^,† Maria Manuela Estevinho, MD,*^,‡ and Fernando Magro,

MD, PhD*^,§

Background: The roles dipeptidyl peptidase 4 (DPP4), aminopeptidase N (APN), and their substrates in autoimmune diseases are being in-

creasingly recognized. However, their significance in inflammatory bowel diseases (IBD) is not entirely understood. This systematic review aims

to discuss the pathophysiological processes related to these ectopeptidases while comparing findings from preclinical and clinical settings.

Methods: This review was conducted according to the PRISMA guidelines. We performed a literature search in PubMed, SCOPUS, and Web

of Science to identify all reports from inception until February 2020. The search included validated animal models of intestinal inflammation

and studies in IBD patients. Quality assessment was performed using SYRCLE’s risk of bias tool and CASP qualitative and cohort checklists.

Results: From the 45 included studies, 36 were performed in animal models and 12 in humans (3 reports included both). Overall, the methodo-

logical quality of preclinical studies was acceptable. In animal models, DPP4 and APN inhibition significantly improved intestinal inflammation.

Glucagon-like peptide (GLP)-1 and GLP-2 analogs and GLP-2-relase-inducing drugs also showed significant benefits in recovery from inflamma-

tory damage. A nonsignificant trend toward disease remission with the GLP-2 analog teduglutide was observed in the sole interventional human

study. All human studies reported an inverse correlation between soluble DPP4/CD26 levels and disease severity, in accordance with the proposal

of DPP4 as a biomarker for IBD.

Conclusions: The use of DPP4 inhibitors and analogs of its substrates has clear benefits in the treatment of experimentally induced intestinal

inflammation. Further research is warranted to validate their potential diagnostic and therapeutic applications in IBD patients.

Key Words: pathogenesis, inflammation, translational, biomarker, ectopeptidase

INTRODUCTION

Inflammatory bowel diseases (IBDs) are a group of

chronic relapsing autoimmune disorders, comprising Crohn’s

disease (CD),¹ ulcerative colitis (UC),² and an intermediate

spectrum of unclassifiable conditions designated as indeter-

minate colitis.³ Inflammatory bowel disease present many

extraintestinal manifestations and may pertain to a cluster

of autoimmune diseases affecting the same patient.⁴ Left un-

treated, these conditions are highly debilitating and potentially

life-threatening and represent a high economic burden.^{1, 2, 4}

In recent decades, IBD has been the target of intensive

research, with considerable progress in the therapeutic man-

agement of the disease and related pathologies. However, most

available treatment strategies have a significant amount of

nonresponders and a wide range of adverse effects.^{1, 2} Thus,

the development of new and optimized therapeutic weapons is

now supported by research on the underlying pathophysiolog-

ical mechanisms of IBD.

Dipeptidyl peptidase 4 (DPP4), homologous to cell-

surface marker CD26 (cluster of differentiation 26), is a

near-ubiquitous membrane protease that cleaves N-terminal

dipeptides from many endogenous and exogenous peptides.⁵

doi: 10.1093/ibd/izaa324

Published online 9 December 2020

Downloaded from https://academic.oup.com/ibdjournal/article/27/7/1153/6028655 by University Library Zurich / Zentralbibliothek Zurich user on 31 December 2022

1154

Melo et al

Inflamm Bowel Dis • Volume 27, Number 7, July 2021

This enzyme is well-known for its physiological action in the

incretin axis as the main cause for the rapid inactivation of the

incretins glucagon-like peptide (GLP)-1 and gastric inhibitory

polypeptide (GIP).⁵ The GLP-1 analogs and DPP4 inhibitors

(DPP4i), which act by increasing the half-life of endogenous

GLP-1, are potent insulinotropic drugs used in the therapy of

diabetes mellitus.^{6, 7}

In the last 2 decades, interest in DPP4, aside from its

utility in glycemic control, has grown. In fact, plasma levels of

this protein have been inversely correlated with disease severity

in IBD and other autoimmune diseases (ADs), making it a po-

tentially new biomarker and a valid therapeutic target for these

conditions.⁸

The 766-amino acid (aa) CD26/DPP4 is a membrane pro-

tein, with many distinct physiological roles (Fig. 1). It contains

an independent C-terminal catalytic region, a cysteine-rich re-

gion, a glycosylation-rich region, a flexible stalk, a transmem-

brane domain, and a short cytosolic tail. It presents specific

binding sites to fibronectin (and other extracellular matrix

components) and extracellular adenosine deaminase (ADA).⁹

After dimerization, CD26/DPP4 is able to activate intracellular

signaling pathways as a type 2 membrane receptor.⁵ The mech-

anism underlying its cleavage and shedding to plasma in its sol-

uble form, sCD26/DPP4 (aa 39–766), is still unclear.¹⁰ Either

through direct cell-signaling or by cleaving immune mediators,

it interferes in several immunoregulatory processes.⁹ By binding

to caveolin-1 on the surface of antigen-presenting dendritic cells

(DCs), DPP4/CD26 stimulates the expression of costimulatory

CD86, through NF-κB signaling, thereby promoting T-cell ac-

tivation.¹¹ Cosignaling with CD45, it enhances T-cell expression

FIGURE 1. Proposed mechanistic view of CD26/DPP4-centered interactions in intestinal inflammation—Th1-polarizing perspective. Membrane

CD26/DPP4, coupled to CD45 in naïve T helper cells, directly stimulates NF-κB-dependent T-cell activation and differentiation into a Th1 phenotype.

Additionally, it proteolytically cleaves N-terminal dipeptides from a variety of substrates, including GLP-1, GLP-2, VIP, and NPY. GLP-1 is a negative

regulator of NF-κB and costimulatory DC signals. GLP-2, acting via GLP-2R, is an intestinotrophic peptide that stimulates the production and release

of intestinal growth factors (EGF, KGFR, IGF-1), which stimulate iSC proliferation and differentiation, effectively counteracting mucosal inflamma-

tory lesions. GLP-23–33 acts as partial agonist/antagonist at GLP-2R. VIP is a negative regulator of neutrophil and lymphocyte chemotaxis and inhibits

macrophage activation. NPY is an ENS-derived pro-inflammatory peptide. CD26/DPP4 is shed to plasma through a mechanism not yet completely

understood. Abbreviations: BNP, brain natriuretic peptide; CCL/CXCL, Chemokines; CCR/CXCR, Chemokine receptors; CD, cluster of differentiation;

DC, dendritic cell; ENS, enteric nervous system; EGF, epidermal growth factor; GHRH, growth hormone-releasing hormone; GIP, gastric inhibitory pol-

ypeptide; GLP-1/-2/-1R/-2R, glucagon-like peptide 1/2/1-receptor/2-receptor; GPR40, G-protein coupled receptor 40 (FFA1, free fatty acid receptor

1); GPR120, G-protein coupled receptor 120 (FFA4, free fatty acid receptor 1); IGF-1/-1R, insulin-like growth factor 1/1-receptor; IL, interleukin; IFN-γ,

interferon-gamma; iSC, intestinal stem cell; iSEMF, intestinal sub-epithelial myofibroblasts; KGF/KGFR, keratinocyte growth factor/receptor; MMP,

matrix metallopeptidase; nNOS, neuronal nitric oxide synthase; NO, nitric oxide; NPY, neuropeptide Y; PACAP, pituitary adenylate cyclase-activating

polypeptide; PYY, peptide YY; sCD26/DPP4, soluble CD26/DPP4; SDF-1/CXCL12, stromal cell-derived factor 1; SP, substance P; TGR5, G-protein

coupled bile acid receptor 1 (GPBAR1); Th, T helper cell; TNF-α, tumor necrosis factor alpha; VIP, vasoactive intestinal peptide; Y1, NPY receptor type

1. Substrates in orange have altered receptor subtype specificity after cleavage by DPP4. Substrates in black mainly lose their bioactivity.

Downloaded from https://academic.oup.com/ibdjournal/article/27/7/1153/6028655 by University Library Zurich / Zentralbibliothek Zurich user on 31 December 2022

1155

Dipeptidyl Peptidase 4 as a Therapeutic Target and Serum Biomarker in IBD

Inflamm Bowel Dis • Volume 27, Number 7, July 2021

of interleukin (IL)-2 and ultimately contributes to differentia-

tion into a T helper (Th) 1 phenotype.^{12, 13} Adenosine deaminase

binding to CD26 colocates ADA to the cell surface, allowing

local degradation of adenosine (a known inhibitor of T-cell

activation) to inosine, hence controlling its extracellular levels

at the immunological synapse. Furthermore, ADA binding to

CD26 was found to produce a direct costimulatory response in

T-cell activation.^{14, 15}

Many bioactive or inactive precursor peptides are sub-

strates of DPP4, including peptide hormones (GLP-1, GLP-

2), neuropeptides (neuropeptide Y, substance P), and many

chemokines and growth factors. Additionally, this enzyme acts

as long as the penultimate amino acid is either proline or ala-

nine. It is also capable of cleaving N-terminal X-glycine/serine/

valine/leucine, albeit at a slower rate. Cleavage is hindered by

the presence of proline in NH2-Xaa-Xaa-Pro position.⁵ In in-

flammatory settings, the most relevant substrates of DPP4 are

GLP-1, GLP-2, and vasoactive intestinal peptide (VIP).

In addition to its insulinotropic action, GLP-1, acting

through GLP-1R, stimulates protein kinase A  (PKA) and

leads to the inhibition of the T-cell costimulatory CD28/

CD86 signal.¹⁶ It is rapidly inactivated by DPP4 into GLP-

19–36/7, with a short half-life of around 1 to 4 minutes.¹⁷ The

GLP-2 is a sister molecule of GLP-1 that is co-expressed and

coreleased from enteroendocrine L cells, after processing of

their common precursor, proglucagon. Through its receptor,

GLP-2R, localized in enteric nervous plexuses (myenteric, sub-

mucosal)¹⁸ and subepithelial myofibroblasts,¹⁹ stimulates the

release of key intestinal growth factors such as KGF, IGF-1,

and EGF; this mechanism reverts inflammatory changes in

the intestinal epithelium.^{20, 21} As with GLP-1, GLP-2 is rap-

idly inactivated by local DPP4 to GLP-23–33, with a half-life of

7 minutes; however its metabolite has a half-life of around 27

minutes.²² Additionally, GLP-23–33 acts as a competitive inhib-

itor of GLP-2 at GLP-2R.²² Teduglutide, a DPP4-resistant long

half-life GLP-2 analog, is approved for use in short bowel syn-

drome and is under investigation for its applicability in IBD.²³

In addition, GLP-2 stimulates the release of VIP from enteric

neurons.²⁴ Vasoactive intestinal peptide is a 28-amino acid pep-

tide with a short half-life of around 1 minute and a wide range

of effects, including neurotransmitter, immunomodulatory,

and secretagogue activities.^{25, 26} It inhibits TNF-α production

by macrophages²⁷ and promotes TH cell differentiation toward

a Th2 phenotype.²⁸

Aminopeptidase N (APN) is an ectopeptidase homolo-

gous to CD13.²⁹ It is being studied in the setting of hemato-

logical disorders and gained interest as a potential co-effector

of DPP4/CD26 in immune regulation.³⁰ Aminopeptidase

N is involved in antigen processing and interaction with ex-

tracellular matrix proteins.³⁰ Its substrates include several

immunoregulatory molecules.³⁰ It cleaves off N-terminal neu-

tral amino acids of oligopeptides but stops if proline is in

the penultimate position.³¹ These catalytic specificities, the

subcellular localization similar to CD26, and increased expres-

sion in activated T cells point to a potential role of APN/CD13

as a DPP4/CD26-substrate generator and vice versa, acting in

tandem as regulators of immune responses. This is further sup-

ported by an improved anti-inflammatory response of the dual

APN/DPP4 inhibition compared with antagonism of only one

of either of these proteins.³² For this reason, we decided to ex-

tend the scope of this review to also include reports on the role

of APN/CD13 in intestinal inflammation.

Although some reviews on the importance of DPP4 and

APN on AIDs and inflammation have been published, the

knowledge on their specific role in IBD pathogenesis is lim-

ited. This systematic review aims to fill this gap and provide a

link between preclinical and clinical data on the role of these

ectopeptidases in IBD (as well as other molecules of their re-

lated axes), through the use of validated animal models of in-

testinal inflammation³³ and studies on IBD patients.

METHODS

This review was conducted following the recommenda-

tions of the Preferred Reporting Items for Systematic Reviews

and Meta-Analyses Statement (PRISMA 2009).³⁴ Study

screening was conducted in 3 electronic databases: PubMed,

Web of Science, and SCOPUS, covering all reports published

through February 4, 2020. The query used for PubMed was as

follows: “DPPR OR DPPIV OR Dipeptidyl peptidase 4 OR

Dipeptidyl peptidase IV OR CD26 OR ADCP2 OR Adenosine

deaminase complexing protein 2 OR Aminopeptidase N OR

Alanyl aminopeptidase OR Alanine aminopeptidase OR CD13

OR AAP OR APN AND inflammatory bowel disease OR

crohn’s disease OR ulcerative colitis.” Similar queries were used

for the other 2 databases, after syntax adaptation. To ensure

the inclusion of all pertinent studies, the reference lists of the

included reports were reviewed by 2 independent researchers.

All preclinical reports emphasizing the pathophysiolog-

ical role of DPP4/CD23 and APN/CD13 in intestinal inflam-

mation models, in addition to clinical reports demonstrating a

link between DPP4/APN and IBD, were assessed for inclusion.

Reports concerning other molecules pertaining to the physio-

logical axes of these factors were also considered, as long as

there was at least an indirect link to DPP4 or APN.

The inclusion criteria were (1) articles studying the asso-

ciation between DPP4/CD26 and IBD in animal models of in-

testinal inflammation and human patients; (2) articles studying

the association between APN/CD13 and IBD in animal models

of intestinal inflammation; and (3) articles studying the associ-

ation between related molecules of the DPP4 axis with known

therapeutic potential, specifically GLP-1 and GLP-2 (and mol-

ecules that influenced GLP-1 or GLP-2 levels, such as TGR5,

G-protein coupled receptor [GPR]-40, and GPR120), and IBD

in animal models of intestinal inflammation and human pa-

tients. No restrictions on publication language were applied.

Downloaded from https://academic.oup.com/ibdjournal/article/27/7/1153/6028655 by University Library Zurich / Zentralbibliothek Zurich user on 31 December 2022

1156

Melo et al

Inflamm Bowel Dis • Volume 27, Number 7, July 2021

The exclusion criteria were (1) review papers, metanalyses,

letters, commentaries, guidelines, editorials, meeting abstracts,

and case reports; (2) studies with no relation to IBD or related

animal models of colitis (intestinal cancer studies, radiation-

induced injury, etc.); (3) studies with no pathophysiological

association with DPP4 or APN; (4) studies including other

substrates of DPP4 and APN (setting a limit to study screening

and to avoid an overreaching and unfocused review of all pos-

sible substrates and their influence in DPP4-dependent inflam-

matory pathways); and (5) studies without available abstract.

The risk of bias in individual studies was assessed using

quality evaluation tools/scales adapted to study type. For pre-

clinical animal studies, SYRCLE’s risk of bias (RoB) tool was

used.³⁵ For studies in humans, CASP checklists were used for

qualitative³⁶ and cohort³⁷ reports. These tools were applied

by 2 independent reviewers, and discrepancies were solved by

consensus.

RESULTS

Study Selection and Characteristics

Study selection, following the PRISMA 2009 Flow

Diagram,³⁴ is outlined in Figure 2. Of the 45 studies selected for

review, 36 used animal models of intestinal inflammation,^38–73

and 12 concerned human IBD patients^{50, 66, 67, 74–82} (3 reports

included both^{50, 66, 67}). The characteristics and main results of

animal studies are described in Supplementary Table 1 and

those of human studies in Table 1. For animal studies, disease

and animal models, with respective variations, are compiled in

Supplementary Table 2.

Most animal studies consisted of interventional proto-

cols of varying durations that tested the effects of DPP4i,

APNi, GLP-1, or GLP-2 analogs (mostly long-acting, DPP4-

resistant, or by continuous infusion) and related drugs on the

recovery from experimentally induced intestinal lesions.

The most used model of colitis was the dex-

tran sulfate sodium (DSS) model (n  =  25), followed by

2,4,6-trinitrobenzenesulfonic acid (TNBS, n = 7), indomethacin

(n = 5), HLA-B27 (n = 2), CD4+ transfer (n = 1), and irinotecan

(n = 1) models. The most commonly used strains of mice were

Balb/c and C57BL/6 mice. Seven reports utilized knock-out

mice for CD26/DPP4, and 1 study used Glp2r^−/− mice.

Despite wide variations in protocol, we decided to inter-

pret each model’s group as a whole, inasmuch as all experiments

were able to elicit an inflammatory response to some extent and

were therefore deemed internally valid.

With a single exception of an interventional study,⁷⁴ all

studies in human IBD patients were observational, focusing

on serum activity, levels of DPP4, APN, GLP-2, GLP-2R, and

others, in addition to associated clinical and endoscopic find-

ings. None of the included studies in IBD patients used DPP4

inhibitors.

FIGURE 2. Study screening and selection process.

Downloaded from https://academic.oup.com/ibdjournal/article/27/7/1153/6028655 by University Library Zurich / Zentralbibliothek Zurich user on 31 December 2022

1157

Dipeptidyl Peptidase 4 as a Therapeutic Target and Serum Biomarker in IBD

Inflamm Bowel Dis • Volume 27, Number 7, July 2021

TABLE 1. Characteristics and Main Results of Human Studies

Authors (Year)

Study

Origin

Study Type IBD

IBD

Assessment

Disease Character-

istics

Intervention

Outcomes Assessed

Results

Buchman AL et al

(2010)

USA and

Canada

(multicenter)

Prospective

(pilot)

CD, moderate-

to-severe (n

= 100)

(Extension: n

= 48)

CDAI

CDAI 220–450

Mean disease duration

11.1–13.7 ± 9.5–10.9

24% concomitant

therapy with

immunomodulators

Teduglutide

0.05/0.10/0.20

mg/kg/d, sc,

1id, 8w

(Extension:

teduglutide

0.10 mg/kg/d,

sc, 1id, for

12 additional

weeks)

Primary: response (100pts

CDAI) or remission (CDAI

≤ 150) at week 8;

Secondary: response and

remission at weeks 2 and

4, mean changes in disease

severity (CDAI), mean de-

crease in # of liquid bowel

movements, mean decrease

in CRP.

No statistically significant dif-

ferences from placebo.

In the 0.2 mg/kg/d group, 40%

achieved remission at 8 weeks.

No difference was observed

in serum CRP at any

timepoint.

Plasma citrulline substantially

increased over time in all

teduglutide groups.

El-Jamal N et al

(2014)

Not

reported

Cross-sec-

tional

CD (n = 19)

UC (n = 15)

Not reported

Insufficient data

—

GLP-2R expression in human

intestine of IBD patients.

Significantly higher expression

of GLP-2R mRNA in the

colon of healthy and IBD

patients, vs in the ileum.

Significantly lower levels of

GLP-2R mRNA in healthy

colon and ileal samples

from CD patients, vs con-

trols; and further significant

reduction in inflammed

samples, vs healthy samples;

Significantly lower levels of

GLP-2R mRNA in healthy

ileal samples from UC

patients, vs controls; and

further significant reduction

in inflammed samples, vs

healthy samples;

Significantly lower levels

of GLP-2R mRNA in

inflammed colon samples

from UC patients, vs con-

trols, but no difference in

healthy colon samples of

UC patients, vs controls.

Hildebrandt M et

al (2001)

Germany,

single

center

Cross-sec-

tional

CD (n = 63)

UC (n = 47)

Controls (n =

28)

CDAI

Rachmilewitz

score (UC)

Mean disease duration

CD: 10.8 ± 8.1 s

years

UC: 11.9 ± 8.5 s years

—

sDPP4 activity and lympho-

cyte CD26/DPP4 expres-

sion in IBD patiens.

Similar percentages of CD2+/

CD26+ cells in peripheral

blood between patients and

healthy controls.

Higher number of CD25+/

CD26+ and CD2+/CD25+

cells in IBD patients.

DPP4 activity was inversely

correlated with disease

activity, orosomucoid con-

centrations and CRP.

Downloaded from https://academic.oup.com/ibdjournal/article/27/7/1153/6028655 by University Library Zurich / Zentralbibliothek Zurich user on 31 December 2022

1158

Melo et al

Inflamm Bowel Dis • Volume 27, Number 7, July 2021

TABLE 1. Continued

Authors (Year)

Study

Origin

Study Type IBD

IBD

Assessment

Disease Character-

istics

Intervention

Outcomes Assessed

Results

Magro D et al

(2017)

Brazil, single

center

Cross-sec-

tional

CD (n = 20)

CDAI

(active: >150;

remission:

≤150)

Montreal

Classifica-

tion

Active CD: n = 10

Remiting CD: n = 10

—

Correlation between serum

levels of LPS and CD26,

and serum levels of CRP,

interleukins, TNF-a and

CDAI.

Significantly higher levels of

serum LPS amd CRP in ac-

tive and inactive CD group,

vs controls.

Significantly lower levels of

IL-1β, IL-6, IL-17 and

CD26 in CD groups, vs

controls.

Nonsignificant higher level of

TNF-α in active CD group,

vs control (p = 0.056).

Negative correlation be-

tween CRP and LPS in CD

groups.

Moran GW et al

(2012)

Not reported Cross-sec-

tional

CD (n = 26)

CDAI

CRP

CDAI = 174.5 ± 14.26

CRP = 20.4 ± 5.4

mg/L

—

DDP4 expression and correla-

tion to underlying intes-

tinal inflammation (CDAI,

CRP).

Significantly lower levels of

plasma and tissue (terminal

ileum) DPP4 in CD pa-

tients, vs controls.

Significant negative correla-

tion between plasma DPP4

and CRP.

Pinto-Lopes P et

al (2020)

Portugal,

multicenter

Prospective

cohorts

CD (n = 101)

UC (n = 94)

Controls (n =

52)

CD: HBI (clin-

ical remis-

sion: ≤4)

UC: pMS

(clinical

remissio: ≤1)

Montreal

Classifica-

tion

Active CD: HBI = 7.6

± 3.2

Remitting CD: HBI =

1.4 ± 1.3

Active UC: pMS = 4.7

± 2.3

Remitting UC: pMS =

0.1 ± 0.3

—

Role of sDPP4 as a biomarker

of IBD activity (potential

in predicting the need for

treatment escalation and

monitoring response to bio-

logical therapy); correlation

between sDPP4 levels with

endoscopic activity and

clinical activity scores.

Patients with active IBD had

significantly higher serum

CRP and FC levels and

lower sDPP4, vs those in

clinical remission.

sDPP4 levels were negatively

correlated with FC, serum

CRP and DAI.

FC was positively correlated

with serum CRP and DAI.

sDPP4 was inversely cor-

related with both disease

activity scores (HBI and

pMS) and endoscopic ac-

tivity groups (stronger in

CD, vs UC).

sDPP4 activity was signifi-

cantly higher in responders

(stronger in UC, vs CD).

Salaga M,

Mokrowiecka

A, Zielinska M

et al (2017)

Not

reported

Cross-sec-

tional

CD (n = 9)

UC (n = 12)

Controls (n = 8)

13 colon biopsy

samples

29 serum sam-

ples

Montreal

Classifica-

tion

CD:

5-ASA (n = 5)

Anti-TNF-a (n = 4)

UC:

5-ASA (n = 12)

—

Expression of GLP-2 and

GLP-2R in the serum and

colon of IBD patiens.

Significant decrease in the ex-

pression of serum GLP-2 in

CD patients.

Significant decrease in the ex-

pression of colon GLP-2R

in UC patients.

Downloaded from https://academic.oup.com/ibdjournal/article/27/7/1153/6028655 by University Library Zurich / Zentralbibliothek Zurich user on 31 December 2022

1159

Dipeptidyl Peptidase 4 as a Therapeutic Target and Serum Biomarker in IBD

Inflamm Bowel Dis • Volume 27, Number 7, July 2021

Authors (Year)

Study

Origin

Study Type IBD

IBD

Assessment

Disease Character-

istics

Intervention

Outcomes Assessed

Results

Salaga M,

Mokrowiecka

A, Jacenik D et

al (2017)

Not reported Cross-sec-

tional

CD (n = 17)

UC (n = 10)

Controls (n

= 8)

35 colon bi-

opsy samples

56 serum sam-

ples

Montreal

Classifica-

tion

CD:

5-ASA (n = 10)

Anti-TNF-a (n = 7)

UC:

5-ASA (n = 10)

—

Characterization of intestinal

tissue and serum expression

levels of APN and NEP in

IBD patients.

Significantly higher expression

of APN mRNA in colonic

tissue of CD patients

(nonsignificant increase in

UC patients).

Nonsignificant increase in the

expression of NEP protein in

colonic tissue of CD patients.

No changes in serum levels.

Schmidt PT et al

(2005)

Sweden,

single

center

Cross-sec-

tional

CD (n = 4)

UC (n = 15)

Controls (n =

10)

Not reported

Active disease:

CD (n = 4)

UC (n = 5)

Chronic/no inflamma-

tion:

UC (n = 10)

CD:

5-ASA (n = 1)

Glucocorticoids (n

= 1)

Metronidazole (n = 1)

UC:

5-ASA (n = 11)

Prednisolone (n = 1)

Azathioprine (n = 1)

Meal stimu-

lation (430

kcal)

Tissue levels and postprandial

secretion of GLP-2 and

PYY in IBD patients.

No significant differences in

tissue content or plasma

concentration after meal

stimulation of GLP-2 and

PYY between IBD patients

and controls.

Tsukahara T et al

(2015)

Japan, single

center

Cross-sec-

tional

CD (n = 16)

Controls (n =

15)

Not reported

Previous therapeutic

exposure:

Anti-TNF-a (n = 4)

Immunomodulator

(n = 2)

Corticosteroids (n

= 0)

—

Expression of GPR40 and

GPR120 in the ileal mucosa

of CD patients and its cor-

relation with inflammatory

parameters.

Intestinal epithelial cells

express GPR40, but rarely

express GPR120, in the

normal ileal mucosa. Boh

were overexpressed in in-

flamed ileal mucosa.

GPR40 and GPR120 are

co-expressed in L cells (signifi-

cant positive correlation).

HBI values significantly

correlated with GPR120 ex-

pression, but not GPR40.

Signficiantly higher levels of

TNF-α.

Both GPR120 and GPR40

expression levels signifi-

cantly correlated with levels

of TNF-α, but not those of

IL-6 or IL-1β.

No differences in protein

and mRNA expression of

proglucagon in CD patients

vs controls.

TABLE 1. Continued

Downloaded from https://academic.oup.com/ibdjournal/article/27/7/1153/6028655 by University Library Zurich / Zentralbibliothek Zurich user on 31 December 2022

1160

Melo et al

Inflamm Bowel Dis • Volume 27, Number 7, July 2021

Authors (Year)

Study

Origin

Study Type IBD

IBD

Assessment

Disease Character-

istics

Intervention

Outcomes Assessed

Results

Varljen J et al

(2005)

Croatia,

single

center

Cross-sec-

tional

CD (n = 38)

UC (n = 24)

CDAI

UC: Truelove

and Witts’

(TW) classi-

fication

Insufficient data

—

Relation between sDPP4

activity with clinical and

inflammatory parameters in

patients with IBD; potential

of sDPP4 activity as a dis-

tinguishing marker between

CD and UC.

sDPP4 activity was signifi-

cantly decreased in both CD

and UC patients, vs con-

trols, although no signficant

differences were found be-

tween the 2 IBD groups.

DPP4 activity inversely correl-

ates with CDAI score in CD

patients, and TW in UC

patients.

Significant difference in sDPP4

activity between male and

female patients with UC.

No correlation between

sDPP4 activity and routine

laboratory parameters in ei-

ther disease, nor in relation

to the location and exten-

sion of pathological lesions.

Xiao Q et al

(2000)

Canada,

single

center

Cross-sec-

tional

GLP-2

CD (n = 39)

CD without

bowel resec-

tion (n = 30)

CD with bowel

resection (n

= 9)

UC (n = 21)

Healthy con-

trols (n = 14)

Immune con-

trols (n = 38)

DPP4

CD without

bowel resec-

tion (n = 1)

CD with bowel

resection (n

= 5)

UC (n = 1)

Healthy con-

trols (n = 6)

Not reported

Mean disease duration

GLP-2

CD without bowel

resection: 4.5 ± 5.1

years

CD with bowel resec-

tion: 12.9 ± 12.7

years

UC: 4.3 ± 5.9 years

DPP4

CD without bowel

resection: 8 years

CD with bowel re-

section: 14.3 ± 3.8

years

UC: 1 year

—

Abnormalities in the levels

and/or molecular forms of

circulating GLP-2 in IBD

patients.

No differences between

plasma levels of total

immunoreactive (IR)-

GLP-2 between normal

healthy and immunocom-

promised control subjects,

and between normal con-

trols and UC patients.

Total plasma IR-GLP-2 levels

were significantly decreased

in CD patients, vs controls.

No differences in total

IR-GLP-2 levels between

CD subgroups according to

disease site, but significant

decrease in total IR-GLP-2 in

CD patients who had a his-

tory of intestinal resection.

Significantly higher bioac-

tive GLP-21–33 levels in IBD

patients (higher ratio of

GLP-21–33 to GLP-23–33).

Significantly lower plasma

DPP4 activity in IBD pa-

tients, vs normal controls.

Abbreviations: CD#, cluster of differentiation #; CDAI, Crohn’s disease activity index; DAI, disease activity index; HB, Harvey-Bradshaw index; IFN-γ, interferon-gamma; LPS, lipopolysaccharide; pMS, partial

Mayo score; Th, T helper cell; TNF-α, tumor necrosis factor alpha; 5-ASA, 5-aminosalicylic acid/mesalazine

TABLE 1. Continued

Downloaded from https://academic.oup.com/ibdjournal/article/27/7/1153/6028655 by University Library Zurich / Zentralbibliothek Zurich user on 31 December 2022

1161

Dipeptidyl Peptidase 4 as a Therapeutic Target and Serum Biomarker in IBD

Inflamm Bowel Dis • Volume 27, Number 7, July 2021

Quality Assessment

The results of the methodological quality evaluation are

summarized in Supplementary Tables 3–5. Overall, preclinical

studies do not report sufficient details to allow for proper risk of

bias assessment, with most items being answered as “unclear,”

according to SYRCLE’s RoB tool³⁵ (Supplementary Table 3).

Unlike randomized clinical trials (RCTs), experimental studies

in animals usually do not implement tools to assess internal va-

lidity and risk of bias in their study design. Consequently, the

generalizability of their findings is compromised due to poor re-

porting.³⁵ In the revised studies, the most affected domains were

inadequate randomization during allocation of study groups

and outcome assessment and lack of blinding/concealment.

For the included reports on human patients, we used 2

separate CASP checklists, according to study design: qualita-

tive³⁶ (Supplementary Table 4) and cohort³⁷ (Supplementary

Table 5). Human qualitative studies showed a better perfor-

mance under appraisal when compared with animal studies due

to better reporting. Nevertheless, some doubts arise (eg, “Can’t

tell”) in the items related to adequate recruitment, representa-

tiveness, and significance of the study population. Furthermore,

the studies revealed a low ability to extrapolate findings to the

general population, most often due to low sample size and high

homogeneity (eg, only female subjects, only recruited from ter-

tiary centers, etc.). As expected, cohort studies had better clas-

sifications, yet the generalizability of the results was also an

important drawback.

DPP4/CD26 and APN/CD13

Animal studies

None of the studies in CD26^−/− mice with experimentally

induced colitis showed significant differences in histological

and clinical scores compared with wild-type strains.^{44, 46–48, 53}

Conversely, Iwaya et al⁵⁶ demonstrated a significant yet tran-

sient improvement of intestinal inflammation in an early phase

of colitis development in DPP4-deficient (F344/Du) rats.

Reports using DPP4 inhibitors showed either partial or

significant improvement of clinical and histological scores and

reduction of MPO activity and pro-inflammatory cytokine

levels.^{42, 43, 51, 52, 55, 61, 65, 67, 71–73}

Bank et al⁴³ studied the effects of combined inhibition of

both DPP4 and APN on colitis attenuation. The dual DPP4/

APN inhibitor IP12.C6 promoted the expression of TGF-β and

FOXP3 compared with separate inhibition and controls. In a

similar manner, dual inhibition of APN and neprilysin (NEP,

CD10) by sialorphin or a sialorphin analogue also significantly

improved colitis, in part through µ- and κ-opioid receptor-

dependent mechanisms.^{57, 66}

Human studies

Serum CD26/DPP4 expression and/or activity was found

to be significantly lower in IBD patients compared with healthy

controls.^{75–77, 80–82} Patients with active disease showed lower

levels than patients in remission, and sCD26/DPP4 levels were

negatively correlated with disease severity and classical inflam-

matory markers, such as C-reactive protein (CRP).^{75, 77, 80, 82} In

addition, Hildebrandt et al⁸² reported significant increases in

CD25⁺/CD26⁺ and CD2⁺/CD25⁺ peripheral blood lymphocytes

in IBD patients vs controls but no differences in the popula-

tion of CD2⁺/CD26⁺ cells. Moran et al⁷⁶ showed that DPP4 ex-

pression was significantly reduced in tissue samples from the

terminal ileum of IBD patients. Still, the authors reported sig-

nificantly higher levels of DPP4 in a Caco-2 cell-based study

after exposure to rhTNF-α.⁷⁶ As with similar reports, they

showed an inverse correlation between serum DPP4 (sDPP4)

levels and CRP; however, such correlation was not found for

CDAI.⁷⁶

In a recent multicentric prospective cohort undertaken by

our study group,⁷⁷ sDPP4 was found to have a strong inverse

correlation with clinical and endoscopic activity. It performed

equally well in postoperative CD patients. Optimal cutoff

points were defined based on receiver operating characteristics

(ROC) curve analysis of sDPP4 and 2 other biomarkers, CRP

and fecal calprotectin (FC). These were used to predict clinical

activity, endoscopic activity, and treatment escalation in both

CD and UC patients. Stratification according to these cutoffs

in a Kaplan-Meier curve showed that after 1 year, 62.2% of

CD patients and 36.3% of UC patients with DPP4 levels below

the cutoff (≤1452 ng/mL and ≤1472 ng/mL, respectively) had

escalated treatment, as opposed to 7.8% of CD and 4.1% of

UC patients with DPP4 levels above the cutoff. The use of 3 si-

multaneous biomarkers proved to have a higher discriminative

power. Eighty percent of the CD patients with 3 positive bio-

markers escalated treatment after 1 year vs 3.3% of the patients

with triple negative biomarkers. Regarding UC, 85.0% of the

patients with 3 positive biomarkers escalated treatment after

1 year vs none with 3 negative biomarkers. All 3 biomarkers

had a similar ability to distinguish between IBD responders and

nonresponders. In a subset of the IBD population with active

disease but without CRP elevation, sDPP4 was able to discrim-

inate endoscopic activity better than FC.

Aminopeptidase N mRNA expression was significantly

higher in colonic tissue of CD patients (nonsignificant increases

in UC patients and in the expression of NEP protein levels in

CD patients), but no changes were detected in serum levels.⁶⁶

Glucagon-like Peptides

Animal studies

Thirteen reports studied the effects of GLP-2, either

by continuous infusion,^{38, 41} long-acting polymer-coupled

(XTEN,³⁹ PEG^{62, 63}) or microsphere-associated⁶⁹ molecules, or

degradation-resistant analogs.^{45, 49, 54, 60, 70} These studies reported

significant improvements in histological and clinical scores, re-

duction in pro-inflammatory cytokine levels and MPO activity,

Downloaded from https://academic.oup.com/ibdjournal/article/27/7/1153/6028655 by University Library Zurich / Zentralbibliothek Zurich user on 31 December 2022

1162

Melo et al

Inflamm Bowel Dis • Volume 27, Number 7, July 2021

and a significant intestinal growth (intestinotrophic) response.

This was assessed by increased crypt cell proliferation and in-

creased survival/reduced apoptotic rates and manifested by

decreased intestinal length reduction after experimental in-

duction of inflammation. A study conducted in Glp2r^−/− mice

demonstrated a functional Paneth cell defect, with reduced

bactericidal activity and reduced expression of intestinotrophic

factors.⁵⁹ Furthermore, similar outcomes were obtained by ac-

tivation of GPR40 (FFAR1, free fatty acid receptor 1⁸³) and

TGR5 (GPBAR1, G protein-coupled bile acid receptor 1⁸⁴) in

enteroendocrine L cells by specific agonists with increased ex-

pression and release of GLP-2.^{58, 64}

A report on GLP-1 nanomedicine revealed a significant

decrease of pro-inflammatory cytokine IL-1β levels and tissue

damage, in addition to a partial attenuation of the diarrheal

phenotype.⁴⁰

Human studies

In accordance with the results in animal models, a pilot

study⁷⁴ of a marketed GLP-2 analog, teduglutide, demon-

strated a trend toward an increased response and remission rate

in IBD patients, although these differences were not statistically

significant. A significant increase in plasma citrulline levels, an

indirect marker of intestinal mucosal mass, was also reported.

Other human studies found lower expression of GLP-2R

and GLP-2 in an inflammatory setting in IBD patients com-

pared with noninflamed sites and healthy controls.^{50, 67, 78, 81}

Schmidt et al⁷⁸ found no differences in tissue or plasma levels of

GLP-2 or peptide YY (PYY) after meal-stimulation between

IBD patients and healthy controls.

A report on CD patients showed overexpression of both

GPR40 and GPR120 (FFAR4, free fatty acid receptor 4)⁸⁵ in

inflamed ileal mucosa and a GPR120-dependent inhibition of

GPR40-induced GLP-2 expression by L cells, promoted by

upregulation of GPR120 by TNF-α.⁷⁹

DISCUSSION

Dipeptidyl peptidase 4/CD26 and its substrates have been

recognized as important mediators of inflammation and immu-

nity. However, data on the efficacy of manipulating the incretin

axis as a treatment modality in IBD lack consistency.^{49, 70, 74}

Despite the success of DPP4 inhibitors as antidiabetic

drugs, the use of DPP4 inhibitors raises special concerns re-

garding the potential short and long-term adverse effects of

inhibiting a molecule with such a broad spectrum of inter-

actions. This is conditioned by the specific cell population

expressing this protein; the local availability, half-life, and bi-

oactivity of its substrates; reaction rates; and substrate genera-

tion by other proteases (such as APN). In this context, different

tissues may present different metabolic signatures resulting

from DPP4 action, depending on the underlying pathophysio-

logical conditions.

Through its enzymatic activity, DPP4 can inactivate sev-

eral inflammatory mediators, such as Mig (CXCL9), IP-10

(CXCL10), and I-TAC (CXCL11), greatly reducing their che-

motactic activity, although without hindering antiangiogenic

activity.⁸⁶ Conversely, cleavage of chemokine LD78β by CD26/

DPP4 significantly enhanced its lymphocyte and monocyte che-

motactic properties.⁸⁷ In addition, DPP4-truncated products

show different receptor interaction²² and selectivity⁸⁸ compared

with their noncleaved precursor peptide. Drugs acting on these

pathways can, therefore, have different net effects. Further re-

search is warranted to assess for significant differences between

truncated or nontruncated forms, with a comprehensive assess-

ment of their regulatory mechanisms and their relevance to the

inflammatory milieu.

In our systematic review, all studies showed at least a

moderate therapeutic benefit of DPP4i in animal models of co-

litis. On the one hand, these results reflect the inhibition of the

catalytic activity of DDP4i over bioactive substrates such as

GLP-2, GLP-1, and VIP, greatly extending their half-lives. This

indirectly inhibits costimulatory signals of T-cell activation and

the production of Th1-polarizing cytokines and chemokines

while promoting intestinal proliferation and tissue recovery

(Fig. 1). On the other hand, a more direct effect of DPP4i action

cannot be excluded becuase CD26^−/− mice did not possess any

inherent resistance to colitis development, nor did they display

an enhanced rate of repair of the damaged mucosal tissue. The

protective effect of DPP4i was only observed in the presence of

DPP4/CD26, suggesting that DPP4i may have unknown mech-

anisms of action besides blocking catalytic activity.^{53, 72, 73}

In line with these findings, all animal studies with GLP-2

and GLP-1 analogs showed a clear benefit (Supplementary

Table 1) regarding intestinal proliferation, preservation of

tissue architecture, and prevention of weight loss in the setting

of induced inflammatory lesions. This is supported by similarly

positive findings in reports using drugs that induce the release

of endogenous GLP-1 and GLP-2, such as GPR40⁵⁸ and TGR5

agonists.⁶⁴

The extrapolation of these data to humans is challenging.

One study⁷⁴ enrolling patients with moderate to severe CD

treated with the GLP-2 analog teduglutide failed to achieve sig-

nificance from placebo, although a trend toward remission was

observed (Table 1). However, this report was limited because it

was based on a pilot study with a relatively small sample size.

Moreover, the study had a high dropout rate, especially due to

adverse effects (up to 31%) and uncomfortable posology (daily

subcutaneous injections).

Lower expression of GLP-2R was reported in IBD pa-

tients vs controls, further significant reduction was reported in

inflamed tissue vs healthy samples, ans a significant decrease

was reported in the expression of serum GLP-2.^{50, 67} One re-

port, however, found no significant differences between IBD

patients and controls in tissue content or plasma concentration

Downloaded from https://academic.oup.com/ibdjournal/article/27/7/1153/6028655 by University Library Zurich / Zentralbibliothek Zurich user on 31 December 2022

1163

Dipeptidyl Peptidase 4 as a Therapeutic Target and Serum Biomarker in IBD

Inflamm Bowel Dis • Volume 27, Number 7, July 2021

of GPL-2 or PYY after meal stimulation.⁷⁸ Wheter this is a

cause or a consequence of the underlying pathological process

remains to be elucidated. Still, the use of GLP-2 analogs and

DPP4i in IBD is a double-edged sword. Despite their benefits,

these drugs promote GLP-2-dependent intestinal tumor cell

proliferation and migratory activity,⁸⁹ which may be of concern

in IBD patients who are already at higher risk of developing co-

lorectal carcinoma.^{1, 2} Further research is needed to assess their

viability and safety as IBD drugs.

Despite their anti-inflammatory effects, DPP4i also

block DPP4-mediated degradation of pro-inflammatory

chemokines,⁸⁶ possibly contributing to the perpetuation of in-

flammatory stimuli. As such, one could argue that DPP4 inhibi-

tion enhances certain pro-inflammatory pathways and leads to

either a new onset or a flare-up of the underlying IBD or other

inflammatory conditions. Nevertheless, a recent metanalysis

did not find an increased risk of IBD in patients under DPP4i

therapy.⁹⁰ This same mechanism can also be beneficial in spe-

cific pathologies. A  recent study by Barreira da Silva et  al⁹¹

showed that DPP4 inhibition prevented CXCL10 truncation

(and inactivation) and enhanced CXCR3-mediated antitumor

immunity and trafficking of T cells into the tumor parenchyma.

Furthermore, DPP4 is cleaved by a not fully understood mech-

anism (eg, shedding), and its soluble form, which is excreted

to plasma, may have yet uncharacterized endocrinological

effects.¹⁰

Aminopeptidase N acts in synergy with DPP4 for the reg-

ulation of immune responses by cleaving peptides preprocessed

by DPP4 or by generating substrates susceptible to cleavage by

DPP4.⁹² Dual inhibition of APN/DPP4 had statistically signif-

icant beneficial effects in animal models of colitis and may be

of clinical significance as a new direct or adjuvant treatment

modality in many pathologies.^{43, 93}

In recent studies, human Th17 cells were implicated in the

pathogenesis of many autoimmune diseases, including IBD⁹⁴; this

represents a step forward from the previous dichotomous Th1/Th2

paradigm. Bengsch et al⁹⁵ demonstrated that CD26⁺⁺ (highly ex-

pressing) cells express typical markers of type 17 differentiation,

even before stimulation, suggesting that CD26⁺⁺ T cells harbor the

Th17 lineage. Moreover, Th1 and Th2 cells were shown to be com-

patible with an intermediate CD26⁺ phenotype, whereas regulatory

CD25⁺CD127⁻FOXP3⁺ IL-10-producing T cells showed an even

lower expression. Patients with active IBD were found to have the

highest frequency of tissue-infiltrating Th17 cells. A strong increase

was observed in inflamed tissue lesions, where 25%–50% of CD26⁺⁺

T cells produced IL-17 upon stimulation, in contrast to peripheral

blood, where only about 5% of CD26⁺⁺ T cells produced IL-17.

These findings suggest an incomplete differentiation of Th17 cells

in peripheral blood due to the lack of sufficient stimulatory signals

found in the proinflammatory cytokine-rich environment of lesion

sites. Thus, a link has been established between CD26 and Th17,

corroborating the preponderant role of CD26 in autoimmunity.

In the dawning of the microbiome age, DPP4 gains even

more interest. A recent proof-of-concept study demonstrated a

DPP4-like activity of gut microbiota,⁹⁶ extending toward un-

charted territories the importance of intestinal microbiome–

host interactions in pathological settings.^{59, 97}

Dipeptidyl peptidase 4 is also a potential serum bio-

marker for many diseases, including cancer and IBD, which have

become an intense target of investigation in recent years.^{77, 98}

As evidenced in recent studies, sDPP4 levels and activity are sig-

nificantly lower in IBD patients vs healthy controls, and sDPP4

correlated negatively with other disease activity markers such

as C-reactive protein, orosomucoid and fecal calprotectin, and

disease activity scores (Harvey-Bradshaw Index [HBI], partial

Mayo Score [pMS], Crohn’s Disease Activity Index [CDAI],

and Truelove and Witts index [TW]) and endoscopic activity

groups.^{75–77, 80, 82} As reported in our previous study,⁷⁷ an impor-

tant benefit of the use of sDPP4 as a biomarker is the ability to

predict the need for treatment escalation from baseline sDPP4

levels at an early stage of the disease process. This enables the

identification of patients who would benefit a priori from a

more aggressive treatment strategy, avoiding exposure to po-

tentially ineffective drugs and the consequent risk of adverse ef-

fects. A biomarker that could direct clinicians to more effective

treatment options (skipping the necessary steps of treatment

escalation with safety) may, at the end of the road, prove to be

more efficient and resource-sparing, and enable a faster con-

trol of the underlying disease. This translates to obvious gains

for the patient in terms of time, expenses, and quality of life.

Higher-powered prospective studies with larger sample sizes are

needed to confirm these findings and prove their usefulness in a

clinical real-world setting.⁷⁷

This review includes 45 studies and is, to the best of our

knowledge, the first attempt to systematize the role of DPP4,

APN, and related substrates in IBD. It also provides, for the

first time, an overview of the integration of the underlying

pathophysiological processes and potential applications in clin-

ical practice. We highlight our translational approach to this

subject (from preclinical animal models of disease to studies in

IBD patients), which allowed to emphasize the existing know-

ledge gaps within and between both settings.

Nevertheless, this systematic review is hindered by some

limitations. Stemming from the inherent limitations of the in-

cluded reports, the overall quality of preclinical experimental

studies was less than desirable. As illustrated in Supplementary

Table 2, protocol variations within the “same” model and

the considerable variability of animal strains and species

do not allow for a reliable comparison between experiments.

Underreporting of the protocol execution and the subjec-

tive nature of the quality assessment tools also limit the in-

ternal validity and the extrapolation of data from animal to

human subjects. Most reports with human populations suf-

fered for having a cross-sectional design (only one study was

Downloaded from https://academic.oup.com/ibdjournal/article/27/7/1153/6028655 by University Library Zurich / Zentralbibliothek Zurich user on 31 December 2022

1164

Melo et al

Inflamm Bowel Dis • Volume 27, Number 7, July 2021

a prospective cohort⁷⁷) or low sample size (Table 1), and only

one⁷⁴ had an interventional approach akin to animal studies.

CONCLUSION

Despite the ubiquitous position and wide spectrum of

interactions of these ectopeptidases intertwining with many

known—yet poorly understood—pathophysiological processes,

their diagnostic and therapeutic benefits may be soon applied

to the growing roster of clinical tools for the management of

IBD. Still, many concerns regarding their potential for stimu-

lating carcinogenesis and immune dysregulation and their via-

bility as biomarkers still need to be evaluated in depth. Further

research is required to achieve the necessary data robustness

to introduce these new applications into clinical practice with

confidence and safety.

SUPPLEMENTARY DATA

Supplementary data is available at Inflammatory Bowel Dis-

eases online.

ACKNOWLEDGEMENTS

The authors thank Paula Pinto, PharmD, PhD (PMA,

Pharmaceutical Medicine Academy) for providing medical

writing and editorial assistance.

REFERENCES

1. Torres  J, Mehandru  S, Colombel  JF, et  al. Crohn’s disease. Lancet.

2017;389:1741–1755.

2. Ordás  I, Eckmann  L, Talamini  M, et  al. Ulcerative colitis. Lancet.

2012;380:1606–1619.

3. Guindi M, Riddell RH. Indeterminate colitis. J Clin Pathol. 2004;57:1233–1244.

4. Greuter T, Vavricka SR. Extraintestinal manifestations in inflammatory bowel

disease - epidemiology, genetics, and pathogenesis. Expert Rev Gastroenterol

Hepatol. 2019;13:307–317.

5. Lambeir AM, Durinx C, Scharpé S, et al. Dipeptidyl-peptidase IV from bench to

bedside: an update on structural properties, functions, and clinical aspects of the

enzyme DPP IV. Crit Rev Clin Lab Sci. 2003;40:209–294.

6. Deacon  CF, Holst  JJ, Carr  RD. Glucagon-like peptide-1: a basis for new ap-

proaches to the management of diabetes. Drugs Today (Barc). 1999;35:159–170.

7. Mest HJ, Mentlein R. Dipeptidyl peptidase inhibitors as new drugs for the treat-

ment of type 2 diabetes. Diabetologia. 2005;48:616–620.

8. Cuchacovich M, Gatica H, Pizzo SV, et al. Characterization of human serum

dipeptidyl peptidase IV (CD26) and analysis of its autoantibodies in patients

with rheumatoid arthritis and other autoimmune diseases. Clin Exp Rheumatol.

2001;19:673–680.

9. Klemann C, Wagner L, Stephan M, et al. Cut to the chase: a review of CD26/

dipeptidyl peptidase-4’s (DPP4) entanglement in the immune system. Clin Exp

Immunol. 2016;185:1–21.

10. Röhrborn D, Eckel J, Sell H. Shedding of dipeptidyl peptidase 4 is mediated by

metalloproteases and up-regulated by hypoxia in human adipocytes and smooth

muscle cells. FEBS Lett. 2014;588:3870–3877.

11. Ohnuma K, Yamochi T, Uchiyama M, et al. CD26 mediates dissociation of Tollip

and IRAK-1 from caveolin-1 and induces upregulation of CD86 on antigen-

presenting cells. Mol Cell Biol. 2005;25:7743–7757.

12. Torimoto  Y, Dang  NH, Vivier  E, et  al. Coassociation of CD26 (dipeptidyl

peptidase IV) with CD45 on the surface of human T lymphocytes. J Immunol.

1991;147:2514–2517.

13. Trowbridge IS, Thomas ML. CD45: an emerging role as a protein tyrosine phos-

phatase required for lymphocyte activation and development. Annu Rev Immunol.

1994;12:85–116.

14. Martín M, Huguet J, Centelles JJ, et al. Expression of ecto-adenosine deaminase

and CD26 in human T cells triggered by the TCR-CD3 complex. Possible role of

adenosine deaminase as costimulatory molecule. J Immunol. 1995;155:4630–4643.

15. Pacheco R, Martinez-Navio JM, Lejeune M, et al. CD26, adenosine deaminase,

and adenosine receptors mediate costimulatory signals in the immunological syn-

apse. Proc Natl Acad Sci U S A. 2005;102:9583–9588.

16. Zhu  T, Wu  XL, Zhang  W, et  al. Glucagon like Peptide-1 (GLP-1) modulates

OVA-induced airway inflammation and mucus secretion involving a protein ki-

nase A (PKA)-dependent nuclear factor-κB (NF-κB) signaling pathway in mice.

Int J Mol Sci. 2015;16:20195–20211.

17. Vahl TP, Paty BW, Fuller BD, et al. Effects of GLP-1-(7-36)NH2, GLP-1-(7-37),

and GLP-1- (9-36)NH2 on intravenous glucose tolerance and glucose-induced

insulin secretion in healthy humans. J Clin Endocrinol Metab. 2003;88:1772–1779.

18. Pedersen J, Pedersen NB, Brix SW, et al. The glucagon-like peptide 2 receptor is

expressed in enteric neurons and not in the epithelium of the intestine. Peptides.

2015;67:20–28.

19. Leen  JL, Izzo  A, Upadhyay  C, et  al. Mechanism of action of glucagon-like

peptide-2 to increase IGF-I mRNA in intestinal subepithelial fibroblasts.

Endocrinology. 2011;152:436–446.

20. Bjerknes M, Cheng H. Modulation of specific intestinal epithelial progenitors by

enteric neurons. Proc Natl Acad Sci U S A. 2001;98:12497–12502.

21. Fesler  Z, Mitova  E, Brubaker  PL. GLP-2, EGF, and the intestinal epithe-

lial IGF-1 receptor interactions in the regulation of crypt cell proliferation.

Endocrinology. 2020;161. doi: 10.1210/endocr/bqaa040.

22. Thulesen J, Knudsen LB, Hartmann B, et al. The truncated metabolite GLP-2

(3-33) interacts with the GLP-2 receptor as a partial agonist. Regul Pept.

2002;103:9–15.

23. Kochar B, Long MD, Shelton E, et al. Safety and efficacy of teduglutide (Gattex)

in patients with Crohn’s disease and need for parenteral support due to short bowel

syndrome-associated intestinal failure. J Clin Gastroenterol. 2017;51:508–511.

24. de  Heuvel  E, Wallace  L, Sharkey  KA, Sigalet  DL. Glucagon-like peptide 2

induces vasoactive intestinal polypeptide expression in enteric neurons via

phophatidylinositol 3-kinase-γ signaling. Am J Physiol Endocrinol Metab.

2012;303:E994–1005.

25. Domschke S, Domschke W, Bloom SR, et al. Vasoactive intestinal peptide in man:

pharmacokinetics, metabolic and circulatory effects. Gut. 1978;19:1049–1053.

26. Iwasaki  M, Akiba  Y, Kaunitz  JD. Recent advances in vasoactive intestinal

peptide physiology and pathophysiology: focus on the gastrointestinal system.

F1000Research. 2019;8:1629.

27. Delgado M, Munoz-Elias EJ, Gomariz RP, et al. Vasoactive intestinal peptide

and pituitary adenylate cyclase-activating polypeptide prevent inducible nitric

oxide synthase transcription in macrophages by inhibiting NF-kappa B and IFN

regulatory factor 1 activation. J Immunol. 1999;162:4685–4696.

28. Delgado M, Gonzalez-Rey E, Ganea D. VIP/PACAP preferentially attract Th2

effectors through differential regulation of chemokine production by dendritic

cells. Faseb J. 2004;18:1453–1455.

29. McClellan JB Jr, Garner CW. Purification and properties of human intestine ala-

nine aminopeptidase. Biochim Biophys Acta. 1980;613:160–167.

30. Ansorge S, Bank U, Heimburg A, et al. Recent insights into the role of dipeptidyl

aminopeptidase IV (DPIV) and aminopeptidase N (APN) families in immune

functions. Clin Chem Lab Med. 2009;47:253–261.

31. Hoffmann T, Faust J, Neubert K, et al. Dipeptidyl peptidase IV (CD 26) and

aminopeptidase N (CD 13) catalyzed hydrolysis of cytokines and peptides with

N-terminal cytokine sequences. FEBS Lett. 1993;336:61–64.

32. Biton A, Bank U, Täger M, et al. Dipeptidyl peptidase IV (DP IV, CD26) and

aminopeptidase N (APN, CD13) as regulators of T cell function and targets

of immunotherapy in CNS inflammation. In: Lendeckel U, Reinhold D, Bank

U, (eds). Dipeptidyl Aminopeptidases. Boston, MA: Springer US; 177–186. doi:

10.1007/0-387-32824-6_19.

33. Jurjus AR, Khoury NN, Reimund JM. Animal models of inflammatory bowel

disease. J Pharmacol Toxicol Methods. 2004;50:81–92.

34. Moher D, Liberati A, Tetzlaff J, et al.; PRISMA Group. Preferred reporting items

for systematic reviews and meta-analyses: the PRISMA statement. Plos Med.

2009;6:e1000097.

35. Hooijmans CR, Rovers MM, de Vries RB, et al. SYRCLE’s risk of bias tool for

animal studies. BMC Med Res Methodol. 2014;14:43.

36. Critical Appraisal Skills Programme (2019). CASP Qualitative Checklist. https://

casp-uk.net/wp-content/uploads/2018/01/CASP-Qualitative-Checklist-2018.pdf.

Accessed June 24, 2020.

37. Critical Appraisal Skills Programme (2019). CASP Cohort Checklist. https://

casp-uk.net/wp-content/uploads/2018/01/CASP-Cohort-Study-Checklist_2018.

pdf. Accessed June 24, 2020.

38. Alavi K, Schwartz MZ, Palazzo JP, et al. Treatment of inflammatory bowel di-

sease in a rodent model with the intestinal growth factor glucagon-like peptide-2.

J Pediatr Surg. 2000;35:847–851.

39. Alters SE, McLaughlin B, Spink B, et al. GLP2-2G-XTEN: a pharmaceutical

protein with improved serum half-life and efficacy in a rat Crohn’s disease model.

Plos One. 2012;7:e50630.

40. Anbazhagan AN, Thaqi M, Priyamvada S, et al. GLP-1 nanomedicine alleviates

gut inflammation. Nanomedicine. 2017;13:659–665.

41. Arthur GL, Schwartz MZ, Kuenzler KA, et al. Glucagonlike peptide-2 analogue:

a possible new approach in the management of inflammatory bowel disease. J

Pediatr Surg. 2004;39:448–452; discussion 448.

42. Ban H, Bamba S, Imaeda H, et al. The DPP-IV inhibitor ER-319711 has a pro-

liferative effect on the colonic epithelium and a minimal effect in the amelioration

of colitis. Oncol Rep. 2011;25:1699–1703.

Downloaded from https://academic.oup.com/ibdjournal/article/27/7/1153/6028655 by University Library Zurich / Zentralbibliothek Zurich user on 31 December 2022

1165

Dipeptidyl Peptidase 4 as a Therapeutic Target and Serum Biomarker in IBD

Inflamm Bowel Dis • Volume 27, Number 7, July 2021

43. Bank U, Heimburg A, Helmuth M, et al. Triggering endogenous immunosup-

pressive mechanisms by combined targeting of Dipeptidyl peptidase IV (DPIV/

CD26) and Aminopeptidase N (APN/ CD13)–a novel approach for the treatment

of inflammatory bowel disease. Int Immunopharmacol. 2006;6:1925–1934.

44. Baticic  L, Detel  D, Kucic  N, et  al. Neuroimmunomodulative properties of

dipeptidyl peptidase IV/CD26 in a TNBS-induced model of colitis in mice. J Cell

Biochem. 2011;112:3322–3333.

45. Boushey RP, Yusta B, Drucker DJ. Glucagon-like peptide 2 decreases mortality

and reduces the severity of indomethacin-induced murine enteritis. Am J Physiol.

1999;277:E937–E947.

46. Buljevic S, Detel D, Pugel EP, et al. The effect of CD26-deficiency on dipeptidyl

peptidase 8 and 9 expression profiles in a mouse model of Crohn’s disease. J Cell

Biochem. 2018;119:6743–6755.

47. Detel D, Buljevic S, Pucar LB, et al. Influence of CD26/dipeptidyl peptidase IV

deficiency on immunophenotypic changes during colitis development and resolu-

tion. J Physiol Biochem. 2016;72:405–419.

48. Detel D, Pugel EP, Pucar LB, et al. Development and resolution of colitis in mice

with target deletion of dipeptidyl peptidase IV. Exp Physiol. 2012;97:486–496.

49. Drucker DJ, Yusta B, Boushey RP, et al. Human [Gly2]GLP-2 reduces the severity of co-

lonic injury in a murine model of experimental colitis. Am J Physiol. 1999;276:G79–G91.

50. El-Jamal N, Erdual E, Neunlist M, et al. Glugacon-like peptide-2: broad receptor

expression, limited therapeutic effect on intestinal inflammation and novel role in

liver regeneration. Am J Physiol Gastrointest Liver Physiol. 2014;307:G274–G285.

51. Elkatary  R, Abdelrahman  K, Hassanin  A, et  al. Effect of different doses of

sitagliptin in treatment of experimentally induced colitis in mice. Br J Pharm Res.

2015;7:140–151.

52. Fujiwara K, Inoue T, Yorifuji N, et al. Combined treatment with dipeptidyl pep-

tidase 4 (DPP4) inhibitor sitagliptin and elemental diets reduced indomethacin-

induced intestinal injury in rats via the increase of mucosal glucagon-like

peptide-2 concentration. J Clin Biochem Nutr. 2015;56:155–162.

53. Geier MS, Tenikoff D, Yazbeck R, et al. Development and resolution of exper-

imental colitis in mice with targeted deletion of dipeptidyl peptidase IV. J Cell

Physiol. 2005;204:687–692.

54. Gu J, Liu J, Huang T, et al. The protective and anti-inflammatory effects of a

modified glucagon-like peptide-2 dimer in inflammatory bowel disease. Biochem

Pharmacol. 2018;155:425–433.

55. Inoue T, Higashiyama M, Kaji I, et al. Dipeptidyl peptidase IV inhibition pre-

vents the formation and promotes the healing of indomethacin-induced intestinal

ulcers in rats. Dig Dis Sci. 2014;59:1286–1295.

56. Iwaya H, Fujii N, Hagio M, et al. Contribution of dipeptidyl peptidase IV to

the severity of dextran sulfate sodium-induced colitis in the early phase. Biosci

Biotechnol Biochem. 2013;77:1461–1466.

57. Kamysz E, Sałaga M, Sobocińska M, et al. Anti-inflammatory effect of novel

analogs of natural enkephalinase inhibitors in a mouse model of experimental

colitis. Future Med Chem. 2016;8:2231–2243.

58. Kato  S, Utsumi  D, Matsumoto  K. G protein-coupled receptor 40 activation

ameliorates dextran sulfate sodium-induced colitis in mice via the upregulation

of glucagon-likepeptide-2. J Pharmacol Sci. 2019;140:144–152.

59. Lee SJ, Lee J, Li KK, et al. Disruption of the murine Glp2r impairs Paneth cell

function and increases susceptibility to small bowel enteritis. Endocrinology.

2012;153:1141–1151.

60. L’Heureux MC, Brubaker PL. Glucagon-like peptide-2 and common therapeutics

in a murine model of ulcerative colitis. J Pharmacol Exp Ther. 2003;306:347–354.

61. Mimura S, Ando T, Ishiguro K, et al. Dipeptidyl peptidase-4 inhibitor anagliptin

facilitates restoration of dextran sulfate sodium-induced colitis. Scand J

Gastroenterol. 2013;48:1152–1159.

62. Qi KK, Lv JJ, Wu J, Xu ZW. Therapeutic effects of different doses of polyeth-

ylene glycosylated porcine glucagon-like peptide-2 on ulcerative colitis in male

rats. BMC Gastroenterol. 2017;17:34.

63. Qi KK, Wu J, Wan J, et al. Purified PEGylated porcine glucagon-like peptide-2

reduces the severity of colonic injury in a murine model of experimental colitis.

Peptides. 2014;52:11–18.

64. Sakanaka T, Inoue T, Yorifuji N, et al. The effects of a TGR5 agonist and a

dipeptidyl peptidase IV inhibitor on dextran sulfate sodium-induced colitis in

mice. J Gastroenterol Hepatol. 2015;30(Suppl 1):60–65.

65. Salaga M, Binienda A, Draczkowski P, et al. Novel peptide inhibitor of dipeptidyl

peptidase IV (Tyr-Pro-D-Ala-NH2) with anti-inflammatory activity in the mouse

models of colitis. Peptides. 2018;108:34–45.

66. Salaga  M, Mokrowiecka  A, Jacenik  D, et  al. Systemic administration of

sialorphin attenuates experimental colitis in mice via interaction with mu and

kappa opioid receptors. J Crohns Colitis. 2017;11:988–998.

67. Salaga  M, Mokrowiecka  A, Zielinska  M, et  al. New peptide inhibitor of

dipeptidyl Peptidase IV, EMDB-1 extends the half-life of GLP-2 and at-

tenuates colitis in mice after topical administration. J Pharmacol Exp Ther.

2017;363:92–103.

68. Schmidt PT, Hartmann B, Bregenholt S, et al. Deficiency of the intestinal growth

factor, glucagon-like peptide 2, in the colon of SCID mice with inflammatory

bowel disease induced by transplantation of CD4+ T cells. Scand J Gastroenterol.

2000;35:522–527.

69. Wu J, Qi K, Xu Z, Wan J. Glucagon-like peptide-2-loaded microspheres as treat-

ment for ulcerative colitis in the murine model. J Microencapsul. 2015;32:598–607.

70. Yang PY, Zou H, Lee C, et al. Stapled, long-acting glucagon-like peptide 2 an-

alog with efficacy in dextran sodium sulfate induced mouse colitis models. J Med

Chem. 2018;61:3218–3223.

71. Yazbeck R. Inhibiting dipeptidyl peptidase activity partially ameliorates colitis in

mice. Front Biosci. 2008;13:6850.

72. Yazbeck R, Howarth GS, Butler RN, et al. Biochemical and histological changes

in the small intestine of mice with dextran sulfate sodium colitis. J Cell Physiol.

2011;226:3219–3224.

73. Yazbeck  R, Sulda  ML, Howarth  GS, et  al. Dipeptidyl peptidase expression

during experimental colitis in mice. Inflamm Bowel Dis. 2010;16:1340–1351.

74. Buchman AL, Katz S, Fang JC, et al.; Teduglutide Study Group. Teduglutide, a

novel mucosally active analog of glucagon-like peptide-2 (GLP-2) for the treat-

ment of moderate to severe Crohn’s disease. Inflamm Bowel Dis. 2010;16:962–973.

75. Magro DO, Kotze PG, Martinez CAR, et al. Changes in serum levels of lipopoly-

saccharides and CD26 in patients with Crohn’s disease. Intest Res. 2017;15:352–357.

76. Moran GW, O’Neill C, Padfield P, et al. Dipeptidyl peptidase-4 expression is re-

duced in Crohn’s disease. Regul Pept. 2012;177:40–45.

77. Pinto-Lopes  P, Afonso  J, Pinto-Lopes  R, et  al. Serum dipeptidyl peptidase 4:

a predictor of disease activity and prognosis in inflammatory bowel disease.

Inflamm Bowel Dis. 2020;XX:1–13.

78. Schmidt PT, Ljung T, Hartmann B, et al. Tissue levels and post-prandial secretion

of the intestinal growth factor, glucagon-like peptide-2, in controls and inflam-

matory bowel disease: comparison with peptide YY. Eur J Gastroenterol Hepatol.

2005;17:207–212.

79. Tsukahara T, Watanabe K, Watanabe T, et al. Tumor necrosis factor α decreases

glucagon-like peptide-2 expression by up-regulating G-protein-coupled receptor

120 in Crohn disease. Am J Pathol. 2015;185:185–196.

80. Varljen J, Sinčić BM, Batičić L, et al. Clinical relevance of the serum dipeptidyl

peptidase IV (DPP IV/CD26) activity in adult patients with Crohn’s disease and

Ulcerative colitis. Croat Chem Acta. 2005;78:427–432.

81. Xiao Q, Boushey RP, Cino M, et al. Circulating levels of glucagon-like peptide-2

in human subjects with inflammatory bowel disease. Am J Physiol Regul Integr

Comp Physiol. 2000;278:R1057–R1063.

82. Hildebrandt  M, Rose  M, Rüter  J, et  al. Dipeptidyl peptidase IV (DP IV,

CD26) in patients with inflammatory bowel disease. Scand J Gastroenterol.

2001;36:1067–1072.

83. Kimura I, Ichimura A, Ohue-Kitano R, et al. Free fatty acid receptors in health

and disease. Physiol Rev. 2020;100:171–210.

84. Guo  C, Chen  WD, Wang  YD. TGR5, not only a metabolic regulator. Front

Physiol. 2016;7:646.

85. Milligan G, Alvarez-Curto E, Hudson BD, et al. FFA4/GPR120: pharmacology

and therapeutic opportunities. Trends Pharmacol Sci. 2017;38:809–821.

86. Proost P, Schutyser E, Menten P, et al. Amino-terminal truncation of CXCR3

agonists impairs receptor signaling and lymphocyte chemotaxis, while preserving

antiangiogenic properties. Blood. 2001;98:3554–3561.

87. Proost P, Menten P, Struyf S, et al. Cleavage by CD26/dipeptidyl peptidase IV

converts the chemokine LD78beta into a most efficient monocyte attractant and

CCR1 agonist. Blood. 2000;96:1674–1680.

88. Grandt D, Schimiczek M, Rascher W, et al. Neuropeptide Y 3-36 is an endoge-

nous ligand selective for Y2 receptors. Regul Pept. 1996;67:33–37.

89. Masur  K, Schwartz  F, Entschladen  F, et  al. DPPIV inhibitors extend GLP-2

mediated tumour promoting effects on intestinal cancer cells. Regul Pept.

2006;137:147–155.

90. Radel JA, Pender DN, Shah SA. Dipeptidyl peptidase-4 inhibitors and inflamma-

tory bowel disease risk: a meta-analysis. Ann Pharmacother. 2019;53:697–704.

91. Barreira da Silva R, Laird ME, Yatim N, et al. Dipeptidylpeptidase 4 inhibition

enhances lymphocyte trafficking, improving both naturally occurring tumor im-

munity and immunotherapy. Nat Immunol. 2015;16:850–858.

92. Lendeckel U, Arndt M, Bukowska A, et al. Synergistic action of DPIV and APN

in the regulation of T cell function. In: Hildebrandt M, Klapp BF, Hoffman T,

Demuth H-U, (eds). Dipeptidyl Aminopeptidases in Health and Disease. Boston,

MA: Springer; 123–131. doi: 10.1007/0-306-47920-6_16.

93. Bank U, Tadje J, Helmuth M, et al. Dipeptidylpeptidase IV (DPIV) and alanyl-

aminopeptidases (AAPs) as a new target complex for treatment of autoimmune

and inflammatory diseases-proof of concept in a mouse model of colitis. Adv Exp

Med Biol. 2006;575:143–153.

94. Miossec P, Korn T, Kuchroo VK. Interleukin-17 and type 17 helper T cells. N

Engl J Med. 2009;361:888–898.

95. Bengsch B, Seigel B, Flecken T, et al. Human Th17 cells express high levels of enzy-

matically active dipeptidylpeptidase IV (CD26). J Immunol. 2012;188:5438–5447.

96. Olivares M, Schüppel V, Hassan AM, et al. The potential role of the dipeptidyl

peptidase-4-like activity from the gut microbiota on the host health. Front

Microbiol. 2018;9:1900.

97. Ryan PM, Patterson E, Carafa I, et al. Metformin and dipeptidyl peptidase-4 in-

hibitor differentially modulate the intestinal microbiota and plasma metabolome

of metabolically dysfunctional mice. Can J Diabetes. 2020;44:146–155.e2.

98. Enz N, Vliegen G, De Meester I, et al. CD26/DPP4 - a potential biomarker and

target for cancer therapy. Pharmacol Ther. 2019;198:135–159.

Downloaded from https://academic.oup.com/ibdjournal/article/27/7/1153/6028655 by University Library Zurich / Zentralbibliothek Zurich user on 31 December 2022